Enrico Moretti on Jobs, Cities, and Innovation
Jun 25 2012

Enrico Moretti of the University of California, Berkeley and the author of the New Geography of Jobs talks to EconTalk host Russ Roberts about the ideas in his book. Moretti traces how the economic success of cities and the workers who live there depends on the education of those workers. Moretti argues that there are spillover effects from educated workers--increased in jobs and wages in the city. He uses changes in the fortunes of Seattle and Albuquerque over the last three decades as an example of how small changes can affect the path of economic development and suggests a strong role for serendipity in determining which cities become hubs for high-tech innovation. The conversation concludes with Moretti making the case for increasing investments in education and research and development.

Paul Romer on Growth, Cities, and the State of Economics
Nobel Laureate Paul Romer of New York University talks with EconTalk host Russ Roberts about the nature of growth, the role of cities in the economy, and the state of economics. Romer also reflects on his time at the World...
Charles Marohn on Strong Towns, Urban Development, and the Future of American Cities
Charles Marohn, President of Strong Towns, talks with EconTalk host Russ Roberts urban development and what makes a strong town. The two discuss how the post-World War II approach to town and city planning has led to debt problems and...
Explore audio transcript, further reading that will help you delve deeper into this week’s episode, and vigorous conversations in the form of our comments section below.


Jun 25 2012 at 2:27pm

Interesting and enjoyable. Confirms other evidence that demand for higher-level skills has been the main driver for the widening wage gap. The discussion about the wrong-headed policies of cities like New Orleans was funny and unfortunately, all too believable. City managers think that organizing farmer’s markets will transform a city into a magnet for innovators and trend-setters because Seattle has them! But the discussion left me hanging; is there anything a New Orleans, or Detroit, or Flint, and dozens of other cities could do to vitalize themselves? Moretti’s only recommendation was more and better education, but how about tax rates, for example? An econtalk on the political economy of big cities would complement this episode nicely.

Brian Rosen
Jun 25 2012 at 2:28pm

I thought this was a great podcast. The one point that I wanted to discuss is my theory as to one factor why you might see productivity gains in successful cities that you don’t get in less successful cities. I have worked in several different cities, currently in Boston, and have found that people that have better access to the key decision makers and innovators do a better job (in general and this is irrespective of education). I believe this has to do with access to knowledge. One subject this podcast has explored is how important education is. What I think you haven’t explored as well is the importance of ongoing, informal, education that you get by being close to innovation and other smart people (though you have certainly alluded to it in passing when explaining why top down solutions are destined to fail).  Employers want to spend as little as possible on education so letting the universities do the training in math etc.. Is usefull. It means less training in house which means you can amortize that cost in the form of higher wages. Once employees are in house training/learning continues to be really important but something often underappreciated by management because informal net works have done a great job of training those around them and I think formal education is less helpful in dynamic work environments. However those spheres of “training” have trouble propagating information outside of the floors they work on and lots of trouble spreading information outside of the cities they work. 

Before I worked in Boston we had a well educated and a very skilled workforce. Over the last 10 years many of the most skilled/innovative jobs where I worked have moved (for a variety of reasons) to Boston, India, Hong Kong, and Japan. What I have seen is stagnation back in operation centers outside of where our most talented workers are. More importantly huge improvements in Boston and other locations that I haven’t seen in Europe, Canada, Manilla, Singapore, etc.. because, as I’ve postulated, at those locations workers don’t have access to the same evolving information (despite valiant efforts to share information). My experience is those areas (and I work with all levels of our Operations departments from mail to processing checks to sales and distribution) that have access provide the less educated/less innovative workers the information  necessary to make better decisions. They better understand every aspect of what is happening and why and so they make less mistakes resulting in higher productivity. Some of that is they can talk after a meeting in more detail about a subject, they can more easily be pulled into informal meetings and they are more likely to go to the bar with the decision makers or people close to decision makers and talk shop. And even when we work hard to train across regions there is some broken telephone at play and thus  creating diminishing returns on training. So even though someone very smart might get most of the details their manager or trainer will often leave something out that prevents them from doing as good a job as someone less educated but closer to the action. 

I don’t have a policy prescription from my experience. It is something I and senior managers debate from time to time. I wanted to add this experience to the debate. Especially since it isn’t always true. That has only been my experience in general and I don’t know if it is true for other companies (though I have certainly been told by friends in other industries that they have similar experiences). And while I think education is important I don’t think universities can solve this problem. If someone left the work force to get more education they would fall further behind. 

Bill Haller
Jun 25 2012 at 4:59pm

The lack of opportunity for high-school grads is lack of basic education at the high school level, not lack of practial technical education. All elementary and secondary students have practical technical skills (With a few hours training my grandchildren could fly a stealth fighter!), they lack basic knowledge of how to read, calculate, write, understand written and verbal ommunication. The absence of these skills is the fundamental failure of our elementary and secondary eduational system.

The post-secondary educational system has many, many problems but the employment of those with a high-school education is not one of them. And more money is not the answer! Higher expecations, more discipline, more basic trades education (outside of craft union control), more local control of eduation (outside off educational-union control), abolish the Department of Education and all the federal money that is used to buy-off teacher unions and school administrators, is part of the answer.

The other part of the answer is an education in virtuous personal behavior, which is a whole different comment.

John Berg
Jun 25 2012 at 6:34pm

After my first reading, I’m sure the podcast will get a re-reading. However I immediately missed any discussion of Patents although I understand the role they played among Microsoft, Google, Amazon, and many others as they moved from place to place. A Patent is property and represents wealth. The US protects property rights.

John Berg

John Keck
Jun 25 2012 at 8:06pm

One of the effects of globalization is the driving down of quality. Plastic isn’t metal. Particle board isn’t solid wood. These new cheaper goods win on price only. They don’t hold up like the old stuff and end up costing more as they need to be replaced more often. Talk to anyone who has moved any particle board furniture. Its trash after one or two moves max.

I assert that much inflation is hidden in the substitution of poorly made goods (and their more frequent replacement cost) for higher quality goods. This is why you are seeing different inflation rates across income groups.

George Balella
Jun 26 2012 at 12:08am

Boy you listen to this exchange about how good globalization has been and I wonder how the professors square all that with the median net worth being at the same level it was 20 years ago??? All the benefits or at least a disproportionate sum of the benefits of globalization and mechanization are going disproportionately to a few on the top. It’s a massive failure of our current form of capitalism that many are comfortable ignoring or dismissing.

Jim Feehely
Jun 26 2012 at 12:19am

Hi Russ,

Good discussion on civic policy, but only from the perspective of conventional neo-classical market economics.

Having said that, Enrico is, in my view, correct in respect of investment in education. But that investment must be in authentic education; i.e. enquiry, analysis and knowledge. Increased investment in training for skills to feed present requirements of business will not do the trick. All training does is entrench a belief in the system as is. And it is the system that is wrong about the hegemoneous commitment to economic growth. We really need to be more innovative than that.

We need massive investment in authentic liberal education to produce the thinkers who can begin to engineer a future, technologically and socially, that provides optimised prosperity and fairer distribution of wealth in a low economic growth circumstances if we are to realise future real quality of life and stop destroying the atmosphere and the environment generally.

In essence, we need innovation in social outcomes, not purely in the pursuit of purely economic outcomes that have not produced the promised economic growth or social equity. Just compare the advances the world made in social equity and in individual wealth in the period between world war 2 and when Nixon dismantled Bretton Woods and unleashed deregulation, globalism and steadily widening gaps between the rich and the poor.

The facts are, as I understand them, that the 30 years following the war produced many multiples of the growth since 1975 and social equity has suffered terribly since 1975 all over the world. It is also the case that in western nations, at least, the period since 1975 has seen a serious decline in authentic education which has been replaced by increasingly technocratic training. Russ, I suggest that is why you are sceptical about more college education.

Jim Feehely.

Russ Roberts
Jun 26 2012 at 6:17am

George Balella,

I’d encourage you to listen to last week’s podcast with Jim Manzi. It was a discussion of the difficulties of attributing causal effects in complex systems. You can’t take a particular result–the path of net wealth– and attribute it to globalization. That would be as unscientific as suggesting the trend in net wealth was cuaused by the preponderance of Democrats in the White House over the last twenty years.

Russ Roberts
Jun 26 2012 at 6:30am

Jim Feehely,

See my comment above to George Balella. A lot of things have happened since 1975. You don’t want t blame them all on Richard Nixon or on the changes you attribute to him–globalization, deregulation, and so on.

The biggest change that has happened since 1975 is the change in family structure caused by an increase in the divorce rate. This is a worldwide phenomenon that makes it difficult to measure what is really going on. Because of the increase in the divorce rate, there has been a much bigger increase in the number of families and households relative to population growth. This distorts measures of progress based on the median.

You argue for an increase in liberal education. We have done that to some extent in the US over the last three or four decades. More people do get computer science degrees today than in 1975. But that number is dwarfed by the numbers majoring in the social sciences. And I would guess that it is dwarfed by a single social science, psychology. Check out the recent EconTalk episode with Alex Tabarrok.

Finally, I don’t want a future that is engineered by really smart people. I’d prefer one that emerges through trial-and-error from the bottom up. I think that method has a better track record, but then again, it is hard to measure the consequences of small changes in complex systems.

Jun 26 2012 at 10:14am


Relevant to this discussion, this link above came up on Slashdot a bit ago, talking about the difficulties that skilled workers sometimes have getting hired in the first place in the IT industry.

The first point is that there’s a disconnect between HR and the IT department, to which the skills being asked of candidates for a job are radically more extensive than the actual experience needed. The second point is the issue of in-house training, wherein tech companies spend a small amount of time trying to train otherwise experienced people to use their systems and instead devote most of their time to poaching talent from other companies that already have that experience. Those two things together make hard to get a jo in the tech industry even when you do have years of experience.

Re: Mr. Roberts:

“I’d encourage you to listen to last week’s podcast with Jim Manzi. It was a discussion of the difficulties of attributing causal effects in complex systems. You can’t take a particular result–the path of net wealth– and attribute it to globalization. That would be as unscientific as suggesting the trend in net wealth was cuaused by the preponderance of Democrats in the White House over the last twenty years.” On that point, I can’t wait for the next discussion of austerity measures in which Mr. Roberts points out that, y’know, we can’t REALLY be sure that Reagan’s policies were what brought us out of a depression…

“Finally, I don’t want a future that is engineered by really smart people. I’d prefer one that emerges through trial-and-error from the bottom up.” And on that point, you’ll find people on both the left and the right who are dismayed by the over-reliance on standardized testing as the new metric of school progress.

George Balella
Jun 26 2012 at 10:23am


The position that nothing can be said or done about complex systems I see as a cop-out. Our banking system needs major reform. It IS planned by those on top of it and it is anything but competitive. It is destroying competitive markets. It plays one city against another and one state against another and one country against another. There are so many brilliant professors like yourself who have made all sorts of rational pragmatic suggestions on how we can improve things. The whole complexity of the banking system exist specifically to undermine the pricing mechanism. You had Anat Admati on explaining how we massively subsidize these banking institutions and how a simple thing like increasing equity requirements would facilitate the banks to start really investing rather than taking advantage of free money form the Fed. But the banks are so powerful it hasn’t happened. Bottom up solutions involve listening to what millions and millions of people want to see done to improve our finance system. Top down control does not just come from government officials. Our banking system is as top down and more destructive than anything our government officials are doing. Here is an eye opening interview on just how corrupt our banking institutions our.


Russ Roberts
Jun 26 2012 at 9:16pm

George Balella,

We partly agree on the banks. What we agree on is that they have too much power and that power has allowed them to make a mess of things at our expense. But that’s because of earlier top down planning by “experts,” economists and those “brilliant professors” and others who justified bailing banks out over the last thirty years. Your faith in top down making things better relies on getting the right people at the top.

I don’t want to rely on that. That’s why I’m a bottom up guy. Bottom up takes advantage of diffused knowledge and decentralized power. I think that’s a better political strategy in a world where even brilliant professors don’t know enough and even when they do, they are prone to political influence.

Jim Feehely
Jun 26 2012 at 9:26pm

Hi Russ,

I agree with you on the ‘bottom up trial and error’ approach. When I argue for a return to liberal education, I am not arguing for expert directed social outcomes. If you review my various posts on your excellent blog, you will see that I consistently identify the ‘expert problem’ as the source of much social policy error and, indeed, economic policy error.

And that point is also related to the problem of complexity, a point on which I commented previously. In my view, we cannot hope to wholly understand the complex systems that influence our lives – social systems, biological systems, ecological systems and economic systems. But the role of the expert, it seems, is to convince us that experts do understand and they do not.

As for changes since 1975, certainly there has been a lot of social change. But there has also been massive changes in international finance, capital and currency flows in the pursuit of a global market, so great that it cannot have failed to have had a huge effect on quality of life and social equity. My point here is that neo-classical economics cannot plausibly cling to its pure market theories as the answer for the future, There is simply too much evidence that conventional economics (i.e. both Keynesian and Hayekian) is just plain wrong. Whilst I accept that markets can be efficient at price setting, markets are not an instrument of social policy. Yet our governments continue to set social policy by almost exclusive reference to market economics, not obvious social need. For example, the world currently makes enough food for all world citizens. But markets do not distribute that food so that there is no hunger. That may be economically ‘efficient’ (although I doubt that), but it creates inhuman outcomes and millions die of hunger. Competition is not the answer to everything. But because of the religious devotion to free market economics (and there has never been a truly free market) cooperation and collaboration is too often seen as ‘unethical’ or ‘anti-competitive’.

We need another way. And, in my view, another way will be forced on us in our lifetimes because the system that has developed since the dismantling of Bretton Woods is seriously broken. There is, for example, no real solution to sovereign debt in Europe and even the USA without massive debt write-offs or massive default. But most economists will not accept that as inevitable. In fact, the system broke irreparably when Bear Sterns, then Lehmans etc failed. But we called that a GFC, not a collapse. The very reason the USA and Europe and others engaged in the abomination of socialising the cost of those criminal failures was the fear of ‘system collapse’. All that did was delay the inevitable.

We need another way that is not so hide bound by free market economic hegemony. That cannot be achieved without an education system truly open to alternatives and not simply a propaganda pump for the economic and social systems that have demonstrably failed to deliver their promises of prosperity and social equity.

John Berg
Jun 27 2012 at 12:21am

If I may offer my shoulders so you two could see a bit further for another “natural experiment,” I would offer another data point.

Delaware Valley USA was an geographical entity of the early 1960s and may well have been the first Silicon Valley. It centered on Philadelphia and I need not mention the many schools of higher education that it contained. But I do have to mention many of the competing entities that revolved around the technologies of computers, software, and electronics: Sperry Rand, Univac, Philco, RCA Computer Div, Sarnoff Labs, Emerson, Burroughs, and several smaller research entities like the Franklin Institute. A motivational impulse was Grace Hopper, COBOL, and, I think, some US Navy money.

John Berg

Jun 27 2012 at 2:35am

I am quite astonished that one argument has not been raised about the spillover effect. If you are company in a traded sector, and the cost of living increases, you need to increase wages to retains the workers, at the expense of profits. If no profit remains, you just disappear. If profits are too low, you may move elsewhere. The ONLY solution to remain in place is to increase productivity. So it is not surprising that the measured productivity (thus computed only on the remaining companies) is increasing. But I would not call it spillover effect, more likely Darwinian effect.

Roger McKinney
Jun 27 2012 at 2:38pm

Very interesting! I was involved in economic development in Oklahoma for 13 years and I watched our state make all of the mistakes mentioned. We wasted millions trying to get businesses like Microsoft to move to Oklahoma, never understanding that the Microsofts of the world are where they are largely because the founders wanted to live near family.

Another good comparison is Tulsa and Bentonville, AR. Tulsa competed with Seattle in the 1980’s as the most beautiful cities in the US. But the oil bust in 1986 nearly destroyed Tulsa. Bentonville went from a sleepy village to a major metropolis (for our region) because of WalMart.

Roger McKinney
Jun 27 2012 at 2:43pm

PS, Oklahoma’s main plan for “development” for the past 50 years has been to spend more on education. At the same time, our biggest expert, outside of oil and wheat, is college students. Educated young people leave the state if there are no jobs. They don’t attract jobs.

Krzysztof Ostaszewski
Jun 27 2012 at 5:47pm

Mr. Moretti presents shocking ideological blindness. If high school educated workers see their wages decline over the last 50 years, how can those high schools have zero responsibility for this situation?

Narasimhan Srinivasan
Jun 28 2012 at 12:04am

Dear Russ,

Another amazing podcast and very illuminating conversation with Prof. Moretti.

Regarding the 2nd of your two quibbles as you put it, why would any body who works as a barber or a waiter would move from say, milwaukee, WI to say Boston, MA to do the same job. Most likely he/she would get almost similar salary, but would have to deal with much higher cost of living.

Regarding Silicon Valley, the fact that William Shockley who was a native of California wanted to move back to the region where he grew up. Given his stature at that time, his decision to start Shockley Semiconductor in Palo Alto is undoubtedly the see for the silicon valley. It is ironic that the region he helped create (even though HP was already there) that eventually came to be called Silicon Valley, even though he was adamant about working with Germanium (which ultimately led the so called traitorous eight to quit his company).


Trent Whitney
Jun 29 2012 at 11:14am


Thoroughly enjoyed the podcast, including the brief discussion of the “wonderful” policies implemented by the St. Louis city and county governments. Saw the exact same things happen post-Wash U in Pittsburgh, too.

Dr. Moretti made some very interesting points throughout the discussion, and I had no issues with them until his closing argument. It seemed to me that his pro-investment-in-education argument was almost the same as the pro-home-ownership argument that we’ve heard for the past 40+ years. Just substitute “home owner” for “student” or “college degreed” and it’s the same argument.

Given how much the cost of a college education has already increased, and how it continues to outpace the rate of inflation, isn’t it somewhat likely that more policies promoting college education will cause an “education bubble” similar to the housing bubble? We’d be artifically diverting even more resources to this sector of the economy. If we accept as true the media reports of myriad college students hopelessly in debt with student loans, we’d only exascerbate this problem by, say, further expanding the student loans program.

In the absence of further government intervention, if there is indeed a huge salary premium for college graduates with certain degrees, then shouldn’t we expect to see private loan companies spring up to finance students seeking those certain degrees? Maybe that is already happening in some circumstances, but it seems that the government student loans program treats each degree equally – somebody studying ancient Peruvian culture who has a very limited field of employers upon graduation pays the same loan rate/has the same terms as somebody studying computer engineering, who could have job offers as early as his/her junior year.

I realize that this discussion was at the end of a podcast already over time, so no complaints about not discussing this issue further. Perhaps an opportunity for another future podcast on college education – you certainly imply that you’ve got a lot more to say on the issue (and I think we’d all find it interesting).

Mirza D
Jun 29 2012 at 11:20am

Hello Russ,

I listened to the podcast and I am not sure if I missed the answer to this question or if your guest simply didn’t answer it. Towards the end of the podcast, your guest states that he controlled for IQ, motivation, and ambition in the study that was being discussed. You asked how he did that and I missed the answer to that question. Following last week’s podcast about the accuracy of experiments in social sciences, I’m curious to understand how he was able to do this.

Lance Gillespie
Jul 4 2012 at 8:41pm

I wonder about the use of higher wages as a proxy for greater productivity when it comes to less educated people in better educated regions.

A few years ago I was in the Silicon Valley area (Los Gatos) and there was a long article in the newspaper about the difficulty of staffing local fast food outlets. Local teens, the usual source of such labor, had no interest in such jobs since they were busy plumping up their CV for college admission. Their parents felt the same way, preferring to give their kids pocket money rather than have them take time away from SAT prep and college essay worthy activities.

The result was that several local franchises found that even offering a 25% premium over minimum wage was not enough to attract workers from a lower income area 30 miles away. They got together to subsidize private bus transportation and that, coupled with the 25% premium, filled their staffing needs.

There is no reason to believe that these workers were faster, better, or more competent now that they were being bussed than they had been when they worked 30 miles away. They were not more productive, they were just paid more to met the market demands for labor. The same may be true in all these high eduction centers – they are also all high cost of living centers and you have to pay more in such places.

Mike Fox
Jul 8 2012 at 3:27pm

Good podcast. Like Lance Gillespie above, I was curious about the productivity metric used to determine the “spillover effect.” Is productivity in the non-traded (confusing term that I assume means unskilled) measured by dollars per transaction? So if a cab ride in Seattle costs more than an equivalent length cab ride in Albuquerque, then the cab drivers in Seattle are more productive than the cab drivers in Albuquerque. That is one reasonable way to measure productivity ($/driver, $/mile, $/cab), but to your point and Lance’s above, it would seem to confuse cost of living increases with spillover effects.

Comments are closed.


About this week's guest:

About ideas and people mentioned in this podcast:Books:


Web pages:

Podcasts and Blogs:



Podcast Episode Highlights
0:36Intro. [Recording date: June 14, 2012.] Russ: Topic is The New Geography of Jobs, the ideas in your book. What has changed over the last three decades? What's new about the geography of jobs? Guest: The last three decades have seen two major global changes that have reshaped the American labor market. One change is the acceleration of globalization. And the second is a variety of technological innovations that have affected the productivity of different workers differently. These two changes are global in scope but they have profoundly different effects on the local economy of different metropolitan areas of the United States. So the globalization of technological progress has increased and strengthened the labor market of some areas but has also weakened the labor market in other areas. This has the result of a redistribution of wealth, jobs, and population across metropolitan areas in our country. Russ: So, what are some of the cities that are doing particularly well and have particularly high education, and what are some that are not doing very well and have low levels of education on average? Guest: The level of education is crucial because for the past three decades the average level of education in the workforce has been the key predictor of the economic success of communities. So, communities with a lot of educated workers, especially a lot of college educated workers, have attracted a lot of good employers and this has resulted in pretty strong economies. For example, the San Francisco Bay area is one, Seattle is another one, Boston. But it's not just a blue state/red state thing. There is Austin in Texas, there is Raleigh-Durham in North Carolina. Parts of Dallas. There are a number of highly educated cities that have been able to attract a very vital and vibrant innovation sector, and this has resulted in job creation for the residents. Russ: I want to make a distinction, which you occasionally mention in the book: When you talk about the importance of college-educated workers, it's particularly important for those in what you call the innovation sector--people in high tech, biotech. So, that would include, obviously, the Bay area in San Francisco, particularly the Silicon Valley, San Jose, Palo Alto, Mountain View area. It would include Austin, Boston; it would include, to some people's surprise, Washington, D.C., which has a fairly strong private sector component in addition to its government sector. So, when you are talking about college educated workers, you are talking about, in particular college educated workers who are working in what you call the innovation sector, right? Guest: That's right. But their presence in a local economy matters not just for jobs in the innovation sector, but also matters for job creation in the rest of the labor market. I argue in the book that the innovation sector has grown a lot and that there has been a shift in the industrial structure of this country away from the production of physical goods, into the production of new innovation and new knowledge and new technology. But the innovation sector remains a small part of the overall labor market; by my account it's about 10% of the overall labor market. The vast majority of our jobs are not in high tech, not in manufacturing; they are in the local service sector. About 70% of our jobs are in positions that we could call local service sector, which is everything from barber to taxi drivers and waiters, doctors, lawyers, nurses. That's common across all industrialized economies. There are basically two sectors in an industrialized economy--the local service sector, which provides services and goods to the residents of an area; and then there is what economies call the traded sector, which includes high tech and manufacturing and agriculture and the construction industry, which compete on the international and national market. Now, although the vast majority of our jobs are in the local service sector, it's really the vitality, the productivity and the degree of innovation in the traded sector that is the key to the prosperity of the local community.
6:26Russ: That's a very deep idea and it's a nice piece of economics, mainly because it's I think not obvious to most people. The point you make in the book is that there's very little--at least right now; it's changing a little bit--there's not that much productivity growth in the provision of local services, right? Explain that, and explain then why the other sector, the traded sector's productivity is so important. Guest: Wages [?] leading of a country, of a city increase when the productivity of its workers increases. It makes sense. Productivity is just the number, the amount of goods or services that you sell, per hour of work. In the local service sector there are occupations in industries that are not very likely to experience large increases in productivity. Think about a barber. It takes pretty much the same amount of labor today as it took 50 years ago. Russ: There's a little bit of technology added to it--they get the clippers, sometimes working. Guest: That's right. It doesn't prove it. I'm not claiming. Russ: But the scissors are better. I interviewed a woman who runs a hair salon--she spends a lot for a scissors; they stay sharper, they probably do a better job. But in general, you can't get a haircut in half the time, or a quarter of the time you used to. It's pretty much the same amount of time. Guest: That's right. When you go to a restaurant, there is some increase of productivity, but it's limited. The amount of labor that a waiter spends to wait your table is not all that different from the amount of labor it took 50 years ago. A taxi driver, a bus driver, a teacher, is the same. By and large. There are some important exceptions, but by and large, the local service sector because of the technology they are using to produce services, tends not to grow its productivity very fast. By contrast, the traded sector, which is made of manufacturing and high tech and agriculture and so on, is experiencing a big increase in productivity. Think about all the new technologies that there are that make productivity higher for workers in the manufacturing sector. Now, those technologies that make workers more productive tend to raise the productivity of the traded sector, and therefore they tend to raise wages in that sector. But that increase in wage doesn't stay just in that sector, because workers can move across sectors. A worker can be either a carpenter or work in a factory. The wage needs to be roughly consistent in the two sectors. And that's why ultimately the increase in productivity in the manufacturing sector benefits the wages of all workers, not just the one in that sector. Russ: So, let me be a little bit skeptical about one piece of that. I think obviously it's true in general. So, for example, to get people to be doctors, you have to pay them enough not to be computer designers. So, wages in doctoring--doctors' wages and salaries--have to be competitive with salaries in the traded sector. And of course doctors themselves can move; they can trade themselves; they can move around. So, in that sense, there's a national labor market. But if I have a low skill and I'm not able to do the jobs in the traded sector, the effect is not as large, correct? Guest: That's right. The effect is larger, the closer the skills that are used in the traded sector to the skills that are used in the non-traded sector. Russ: Well said. Guest: That said, if you have a high school degree and you are working as a carpenter, but you could also work potentially as a blue collar worker in a factory. When the productivity in the factory increases, then your wage will at least to some extent account for the increase in the wage in the factory. Otherwise you will just leave construction and move to the factory sector. So, competition among construction firms, they are forced by competition to pay the wage that would get them out of the factory.
11:31Russ: So, you tell a very nice story, which I think is underappreciated, about the transformation of the manufacturing sector, which used to be the driver of productivity growth over a long period of time. Part of the driver, in the middle and early part of the 20th century, and how that's less so today because of productivity, globalization. And that we still make a lot of stuff. You make that point a lot of times; it's very important to remind people that the manufacturing sector in the United States is extraordinarily productive. It just doesn't have as many workers as it used to have. Guest: That's right. I don't think people realize that. But today we are producing, in terms of value, twice as much as we were producing 30 years ago in the American manufacturing sector. So, American manufacturers are doing fine. But they are using fewer and fewer workers. And the reason is that machines effectively substituted for labor, in this case. If you have ever visited a factory, you immediately see it--there is nobody around. Russ: It's incredible. Guest: Everything is made by machines. Russ: The statistic you quote that in the 1950s the average worker at General Motors (GM) could make 7 cars in a year, and now they make 28--that's an unbelievable transformation. Guest: That's right. It means that for the same amount of cars sold, now GM needs 70% fewer workers. That alone means that there are fewer and fewer jobs in this sector. In some sense, this sector becomes so productive and so successful at using the best possible machinery and robots that it needs fewer and fewer workers. Russ: And then I think most people say: What's going to replace all those jobs? And the answer, of course, depends on your skill level. But what has replaced those jobs seems to be to some extent the innovation sector, and you point out that although people think outsourcing is taking away all of those jobs, that's not true. Guest: That's right. I think this is one of the great misconceptions that are out there in the public debate, which is globalization and outsourcing are reducing the demand for workers in our high tech sector and our innovation sector. But the reality is that's not true. Globalization and outsourcing are actually good for labor demand, for the demand for high-skilled workers in our country. And there are two reasons. First of all, outsourcing allows high tech companies in our country to be more competitive and therefore to try better and therefore ultimately to hire more in this country. There are a series of studies on the pharmaceutical industry, the biotech industry, and the high tech industry, and most of them conclude that outsourcing actually helped American companies to hire many skilled workers in our country. The second reason is more general. If you think about globalization, it is an increase in the size of the market that companies have access to. Now, this increase in the size of the market is good in general for all companies, but it is particularly good for companies in the high tech sector and in the innovation sector in general. And the reason I don't think it is fully appreciated is the fact that the cost structure in this sector is such that most of the cost is fixed. If you are in a high tech company, most of the cost you are facing is fixed Research and Development (R&D) costs. Microsoft, most of the cost you have in developing Windows is fixed. You have to pay thousands of hours of work for your software engineers. That cost is fixed. It doesn't depend on how many copies of Windows you sell, because the cost of making an additional copy of Windows is trivial. It's the cost of the CD-ROM. Russ: And we're getting rid of that. Guest: Yeah, exactly. Russ: Now it's just a little bit of bandwidth. Guest: Exactly. So, the variable cost in this industry is very low. And the same is true for a biotech firm. Most of the cost is up front, R&D cost. The cost of making the pill at the end is trivial relative to the cost of all of the scientists in the labs to develop the patent. So, the innovation sector is characterized by fixed costs, large fixed costs and low variable costs. Now for this industry, increasing the size of the market actually is a boon, because it means that your profits increase but your costs don't increase. Russ: As much, very much, yeah. Guest: So this is the reason why I think that globalization is one of the key drivers of the rise of the innovation sector. Russ: As you point out though, when people debate these kinds of issues, they look at one category, say, of high tech workers. They'll look at, say, computer programmers, and they'll see that many companies are outsourcing their computer programming to India or elsewhere. But they don't look at the entire sector. The entire sector has exploded in growth over the last decade. Hugh increases. Guest: That's right. If you look at employment, for example, in software, it has increased much faster than employment in the rest of the labor market over the past 20 years. If you look at employment in the Internet, it has increased 200 times faster than employment in the rest of the labor force. These are sectors that are heavily open to outsourcing, but they are creating more and more jobs in the United States.
17:54Russ: We talked, as I alluded to a minute ago, although many jobs are increasing, those jobs that are increasing are typically skill-related. So, what you hear people complain about, and it's an important concern, is that in the 1950s, if you just finished high school, you could get a good job; and as those jobs have become less numerous because of the productivity changes you are talking about, now if you just graduate high school you can't get a very good job. And as you point out, the wages of high school graduates who do not go on to college, not only are they not growing as quickly as other parts of the labor force, they are actually falling in real terms. Guest: That's right. The average wage for a male, full time worker with a high school education today is about 20% lower relative to 1980. Russ: There are some problems with that, some of which you point out, which are that the deflating Consumer Price Index (CPI) that you use, maybe is not so accurate; and it's even less accurate for low income workers than it is for higher income workers. It doesn't include benefits. But they are not doing so well. That part is clearly true. They are not doing as well as others in the economy. Guest: I think it's unambiguous. I think for the first time in the history of America, the typical worker, the median worker, which is someone with a high school degree today, his wage hasn't increased relative to the previous generation. In fact, it has declined. This reflects, again, these two forces--globalization and technological progress--that have effectively reduced the demand for his skills relative to the demand for other workers' skills. Russ: I think the empirical evidence is a little more ambiguous. I think, besides the issue of the CPI and how you deflate nominal wages to get a real wage, besides the fact that you are not including benefits, you've also got issues of immigration and other factors that have changed who the median is. But I'll accept the point that the growth in productivity and in wages for higher income workers like you and me is much greater. I always think about somebody in 1950 saying: Well, in 1900 you could be a blacksmith and make a good living, and now you can't in 1950. So, that's sort of the situation we are in today. In 1950 you could have made a good living as a factory worker. You can't today. There aren't very many of those jobs to go around. The question is: What should we do about it? Or what should somebody do about it--I don't know if it's "we." What should workers and students and others, what kind of decisions should people make given that that's a reality. There's no doubt, as you point out and as many others have pointed out, the returns to education are very high. So, what should be done? Guest: Well, there are a lot of people who claim that we should defend and protect the American manufacturing sector from this type of threat. Their policy prescriptions are often a combination of subsidies of manufacturing or specific parts of manufacturing, and/or types of protectionist policies that would limit the import of shipped goods from other countries. I am skeptical of these policies. I think they would be counterproductive. And most importantly, I doubt that they would slow the decline in blue collar jobs. The decline in blue collar jobs, as we argued, is driven by structural forces in the economy. And I think it's unlikely to slow down. And it's not just that there are fewer and fewer jobs in manufacturing, but the fact the type of jobs in manufacturing are also less and less the type of standard blue collar job that you had in mind. Russ: That's a great point. Guest: We have lost an average of 360,000 blue collar jobs in manufacturing per year for the past 30 years; but in the same period, the number of engineers in manufacturing has doubled. So, I don't think, if we have in mind protecting jobs for less educated Americans, the solution is to try to throw money at the manufacturing sector and hope for the fact that these blue collar jobs will come back. Russ: As you point out, if you go into a modern factory--I have, and I know you have--as you say, one of the most striking things about it is there's nobody in it. A lot of the people who are in it are there just to make sure that the machines are going. So, that's unlikely to be a great-paying job. The people who can fix those machines, who can customize those machines--interested listeners can go back to the podcast we did with Adam Davidson--there are parts of the manufacturing sector where there is some high skilled work going on. But it's a small sector. It's just not likely to be a place of job growth in general. One of the surprising factors that I found out in researching this book is that if you look at employment in computer manufacturing or semi-conductor manufacturing, those peaked respectively 25 and 12 years ago; and they have been declining ever since. And again, this reflects the fact that even in these advanced productions, when you are making computers or semi-conductors, the twin forces of globalization and automation have reduced the need for bodies in those factories, for blue collar workers. But at the same time the number of highly skilled, highly educated workers in those two industries have increased. And I think the best example of this is Apple. If you have an iPhone or iPad, that iPhone or iPad was designed, engineered, marketed from Cupertino, here in the Valley, by highly skilled American workers. But its physical components, including the very sophisticated electronic component, were by and large not made in America. They were made in Asia--in Singapore and Taiwan. And then finally the product was assembled in Shenzhen, China. So, this type of supply chain, where the innovations part is in America but the physical manufacturing part is outsourced, has become the standard for most American manufacturers. Both in high tech, and in more traditional parts. This means that in these industries, the jobs that remain are highly educated, highly skilled, but they don't have a lot to do with the physical production of things
26:08Russ: Before we leave this topic--and I want to move on to cities, which we'll get back to in a second--you have a section entitled "How China and Wal-Mart Help the Poor." That's an unusual section to find in a book. Talk briefly about that. Guest: Well, that section is motivated by the fact that, again, a lot of people look at globalization and job losses that evolve and they conclude that we would be better off without it. But although, as I said, there are profound job losses for some groups of workers due to outsourcing and globalization of [?] skilled positions, there are also gains in terms of cheaper products. And that section talks about some interesting research that has been done on how much people paid for what they consume. And researchers basically follow a representative sample of households. And these households, every time when they went out shopping and they came back, they were given a scanning machine, and they were scanning the goods that they bought, and they were talking about the price they paid. And what emerges at the end of this study is that different households face different kinds of price changes, different kinds of inflation rates effectively, depending upon their income. Households with low income, they tend to experience less inflation over the past 20 years than households with high income. And this reflects the fact that households with low income are more likely to buy from places like Wal-Mart, and also more likely to buy cheap consumer products that are made cheap by globalization, effectively by the fact that they are made by China or in Vietnam. By contrast, high income households spend a larger fraction of their income on personal services. They are less exposed to global competition, and therefore they experience steeper price increases. Effectively it's as if we have not just one inflation rate in our country. We have multiple inflation rates, depending on the type of goods that you buy. And it turns out, over the past 20 years, thanks to globalization and in part thanks to Walmart, those who have low income have experienced lower inflation rates than those who have high income. Russ: My Dad has been telling me for decades that the CPI is meaningless. I think that's a little strong, but he's onto something. His argument is always: Well, I don't buy what's in the basket; I don't buy the representative basket. And as an economist I always used to argue with him. But as I've become more Austrian in my old age, I'm more interested in the aggregation problems. Guest: Well, the CPI is an average across many, many, many households. Russ: But it's actually worse than that. I think it's an average across many, many retail outlets, and I think the Bureau of Labor Statistics (BLS) has failed to capture the massive volume that Wal-Mart has been able to drive through its stores. And so they continue to sample--for a long time they didn't sample Walmart at all, for example on the food side, I think; it took them a long time to sample Wal-Mart's groceries; maybe they still don't. But they also don't get the weights right, because they don't correct for the volume of the overall economy. I don't know what it actually is, but it's a very important phenomenon. Guest: But even using their own data, what you find is that the CPI is different depending on who you are and depending on your income, because the type of goods that you consumer are different. And if you are low income, you tend to spend proportionally more on goods that have experienced less price increase, exactly thanks to China and Vietnam. And Wal-Mart. Russ: And the magnitude is shockingly large, of this effect. It's not 5% difference, right? Guest: Yes, yes. It's larger. And it's also made larger, by my study on the difference in the cost of living in housing costs. But we'll get to that.
30:39Russ: Let's get to that. Let's move on to cities. You contrast the difference between Seattle and Albuquerque. And one of the reasons it's such a great contrast is we have such a romantic vision of Seattle today, because of Amazon and Microsoft and other high tech companies. But it wasn't always that way. Guest: That's right. As I was searching the book, I was searching for a tale of two cities that could illustrate how the increase in the jobs in the innovation sector can help some cities to grow. And I found what I think is the perfect example in the story of Seattle. As you pointed out, we have this great image of Seattle today--it is vibrant. Russ: Hip. Guest: Cosmopolitan local economy. But in the late 1970s, Seattle was in very different conditions. Seattle as an economy was heavily focused on traditional manufacturing, and services for lumbers and [?] industries. As you can imagine, in the 1970s, these were not great industries to have. The only innovative part of their economy was Boeing. But Boeing was struggling in the late 1970s, laying off people by the thousands. So the economy was in really poor shape, and people were leaving the city by the thousands. At some point the situation was so dire that a billboard went up on the freeway that goes from the airport to downtown that said: The last one out of Seattle, please turn the lights off. And that billboard, to me, was a sign of a community in decline. Russ: It got so bad even their basketball team went to Oklahoma City. But I'm not sure that's related to the story. But go ahead. Guest: But there's something that happened that changed the history of the city forever, and it has to do with Microsoft. Microsoft was not founded in Seattle. At the time it was in Albuquerque, New Mexico. Albuquerque at the time was more high tech than Seattle. And in fact the main reason Microsoft was there was that their first client was there. Microsoft stayed there for four years and was doing fine, was prospering. But in 1979, Bill Gates and Paul Allen, the founders, decided that they wanted to be close to their families--who were in Seattle. So they relocated the company back to Seattle. Now, at the time this was a small company, 15 employees. And nobody really paid attention to this move. But in retrospect, that was the seed that was responsible for the growth of the high tech sector in Seattle and the complete reshaping of the local economy, and the rebirth of the city both economically, culturally, and in terms of amenities. If you look, for example, at wages in Seattle and Albuquerque, you see that at the time of the move they are not all that different, and more important, they are trending in very similar wages; they are changing from year to year in very similar ways. But after the move, as Microsoft started growing and started attracting around it many high tech firms, you start seeing the wage in Seattle growing much faster than Albuquerque. Albuquerque's economy starts struggling, effectively for three decades after this move. But the Seattle economy starts booming. And the larger Microsoft becomes, the larger the gap between the two cities becomes. And it's not just the wage. It's the number of college graduates, for example, in the two cities. Every passing year, the gap increases. It's now almost double in Seattle relative to Albuquerque. Although the two cities were not all that different economically in 1979, they are almost like two different countries. One is an advanced, driven economy; the other one is the struggling middle-income country. Russ: What I loved in your discussion of these two cities, and in economic development in general, is pointing out some of the incorrect lessons that people have learned from this tale of two cities. So, I used to teach at Washington University in St. Louis. The city struggles, very stagnant; the metropolitan area is somewhat healthy, but the city itself has done very badly over the last few decades. I always thought: we need to be more like Seattle. So, they would send city officials to Seattle, and then they would come back and say, well, the U. of Washington is a key part of Seattle's success; they have a lot of great high tech departments; and Washington U. could also be such an anchor. That's somewhat interesting; but as you point out, a lot of great research universities are not--the cities around them are not thriving. But then they'd say, the second thing: Well, Seattle has a lot of farmers' markets, so we'll have a lot of farmers' markets. Maybe they should have made it rain more, try to seed the clouds. And they'd forget things like St. Louis is farther from Mt. Ranier and great skiing and a thousand other things. But your point, which is more important, is it's not just that it's hard to imitate successful cities, and copy them and grow like they did. Your point is that they confuse often what came first and what came second. So, the U. of Washington was a great research university when Seattle was more like Albuquerque. It didn't help. Guest: Yes. That line of reasoning confuses cause with effect, effectively. The U. of Washington in Seattle is now thriving in great part because of donations from the Gates family, the Allen Foundation, and all the wealth that is being created in the city. In 1979 it was certainly not as good as Washington University in St. Louis. And yet Seattle has a high tech cluster and St. Louis doesn't. This tendency of rationalizing ex post based on what you see now is very pervasive. And it's very dangerous, because I think it leads communities to spend vast amounts of money into policies that are not likely to change the local economy. I think the example of the farmer's market is even more pointed. Twenty years ago, 15 years ago, this notion of revitalizing cities using farmers' markets and public [?], other types of local amenities which attract these so-called creative class became very popular thanks to a series of books. But again, there is no solid evidence that suggests that those things have any effect on jobs. They are certainly important for how nice the community is; and I'm not saying we should not invest in these things. But we should do it for the right reasons. If we decide to have farmers' markets, we should do it because it's nice to have fresh fruit, but not because it's going to bring us a high tech sector. Russ: Or because it's going to bring us all these college students who like fresh fruit.
38:47Russ: Talk about Berlin, because that's a great example. Guest: Well, these theories about how important it is to attract the so-called creative class to revitalize a city would predict that cities that have a lot of highly creative young people are the cities that are going to thrive. And I think a great counterexample is Berlin. Berlin, since the fall of the Wall, has become a magnet for highly creative, highly artistic types from all over Europe. It's in part because it's such an interesting, culturally vibrant and politically open type of city. They have a lot of Italians, Spanish, and people with a lot of education--they move to Berlin. It's, according to most guidebooks, is one of the coolest cities in Europe, possibly in the world. But there is one problem with this picture--which is that it doesn't really create many private-sector jobs. Berlin has for the past 20 years had one of the highest unemployment rates in Germany; one of the lowest per capita income among German states. And it basically survives because it's the capital, and therefore it's directly supported by taxpayers everywhere in Germany. And also because it's such a hot tourist destination due to its very peculiar history. But if you look at other sectors--solid, private sector, innovative employers--they have a couple of high tech types, but it's nothing relative to the size of the city. So, to me, the notion of revitalizing St. Louis making it cooler and sexier, it's unlikely to work. If it doesn't work for Berlin, I don't see how it could work for St. Louis or Detroit. Russ: Yeah, I'm with you there. And the other thing they would do is they've created these private-public sector cooperative incubators and other things. Same mistake: There's a lot of startups in Seattle, so we need more startups. This is an emergent phenomenon; it's not something you can steer and direct easily. Stupid mistake. Guest: If you look at the history of America's great innovation hubs, they haven't found one that was directly, explicitly engineered by an explicit policy on the part of the government. It's really hard. This is not how innovation hubs and clusters get developed. They often get developed because of idiosyncratic factors like a local firm succeeds and it starts attracting more firms like that. And this creates a cluster that then becomes stronger and stronger, and that feeds on itself. I think the history of Seattle is a good example. How did Microsoft change the Seattle metro area? Well, Microsoft has 40,000 employees in Seattle. You may say: That's a lot. But the reality is this is a metro area of 2 million people. So, one company alone cannot do it. The real reason why Seattle reshaped the local economy--actually, there are two reasons. First of all is that by being there, Seattle attracted many more high tech and innovative companies there. It became a magnet for the cluster. It became the seed for the cluster. And one great example is Amazon. Amazon was founded in 1995 by Jeff Bezos. At the time, Bezos was not living in Seattle. He had no personal reason to be in Seattle. He was wasn't born in Seattle. In fact, he was born in Albuquerque. When he decided to found Amazon, he places it in Seattle because by 1995, so 15 years after Bill Gates' regional move, Seattle has become a magnet for this type of innovative companies. And now there are tens of thousands of jobs that are there because of Bezos's decision. So, it's not like Microsoft directly helped Jeff Bezos to locate there, but indirectly created the ecosystem that made it almost impossible for someone like Jeff Bezos to pass on Seattle. Russ: And when you look at it at a point in time it looks like these advantages of clustering and hubs are going to persist forever. And then you think about Detroit, which used to be a hub, a tremendous hub of manufacturing activity because of the auto industry. Things change. Guest: Yes. There is recent research that shows that the rise of Detroit as an innovation hub parallels tightly the later rise of Silicon Valley, in terms of job creation, patents, startups, and so on. So, as you pointed out, Detroit used to be the Silicon Valley of its time. It used to have one of the most successful industrial classes of its time. It had some of the most innovative companies, some of the best engineers. And yet it didn't last. And I think that's one of the lessons that places like Silicon Valley should take. What was the failure of Detroit? Well, the failure of Detroit--sellers maul[?] failures. Detroit's brought[?] its local politicians, the unions, and the car executives--they became complacent. And they had their share of responsibility. But the biggest failure of all was the fact that, at the time when it had one of the biggest, most prosperous ecosystems in the world, it failed to reinvent itself and leverage that consistent into something new. When the demand for workers in the auto industry started collapsing, Detroit failed to find something new. The technological frontier keeps evolving, and there's no technology that stays on the frontier forever. The secret for a community is to reinvent itself into new things. And I think this might be the most important difference between Detroit and Silicon Valley. Silicon Valley keeps reinventing itself. It used to be mostly hardware and silicon doctors in the 1970s and 1980s. Then it became mostly software. And then it became mostly internet. And now it's branching out into new things like nanotech and biotech. Nobody knows whether those technologies will succeed and will create jobs, but this is an ecosystem that keeps following what the frontier of technology is. In contrast, Detroit failed to, and now it's probably too late, because the ecosystem is not there any more. And so there is little left to be leveraged.
46:45Russ: Just a small quibble with your choice of language: you are using Detroit and Silicon Valley as if they are purposive decision makers. Of course they are just an agglomeration of individuals. And I think what your example points to is that it's the people in Silicon Valley who have reinvented themselves; and the people in Detroit weren't as able to. But even that somewhat misleads because--and this will be a nice segue into other points in your book--is it really a tragedy that it is what it is? A lot of people have left there, moved to New York City, they've moved to California, to Seattle. The city itself--we can have a romantic association with certain cities, but we really care about the people. Guest: We do. Russ: And the people--so some cities thrive and some don't. And some places rise and some fall. Guest: It has always been happening through the history of America. As I was saying, as the technological frontier changes, some places grow and other places shrink. It used to be that agriculture was the engine of jobs and prosperity, and then when it shifted to manufacturing a lot of the world areas started losing population and so people relocated. I take your point and I agree with that. But if you look at, today, for example, with the divergent fortunes of American cities as large as ever, so we differ in such things such as wages as large as ever, you see that there are vast difference in the ability--there are large differences in the economic forces of different cities, which suggest there are large differences in the economic return to geographic mobility are high. And yet there are vast differences in the ability of relocating and moving to different cities among American workers. Some American workers are much more able to do so, and much more able to, for example, leave declining clusters and move to new cities; and others are much less able. And this is costing the less mobile increasingly in terms of jobs, careers, and incomes.
49:16Russ: Let's talk about that, because that's a key theme in the book, and I must admit I'm a little bit skeptical of it, these spillover effects. The question is, if you have low skill and you live in Detroit, you don't make a lot of money. If you have low skill and you live in Seattle, you make a little bit more. Maybe a lot more, even if you are low-skilled. And your argument is that there are these spillover effects from the higher educated people in a community that benefit the low-skilled workers and that aren't then spilling over in places like Detroit that have not been able to attract high skilled workers. Talk about why you think that's true, and why it's important. Guest: Well, I think it's probably the most important part of the story, because not every worker in America can work for Microsoft or Google or Amazon. As we are saying, the vast majority of our jobs are for people outside high tech. But every time a community is able to attract one of these highly skilled, innovative jobs by attracting a high tech company, it gains not just a job. It gains many more jobs in the local service sector. I call this the multiplier effect. And this multiplier effect is remarkably large. My estimate suggests that for each high tech job that you create in a community, 5 additional jobs are supported in the same community outside high tech in the local service sectors. And this includes a variety of positions, some of which require college degrees, so this would be the doctors, the lawyers, the teachers, and the nurses; but others don't require a college degree. These would be the taxi drivers, the bartenders and the waiters, and the childcare workers. All these jobs are ultimately supported by the fact that there is more wealth now in the community thanks to that new high tech company. So, just to give you a sense of the magnitude here: there is a lot of talk about Apple and Apple job creation. Well, Apple here in Cupertino has 13,000 employees. Mostly as we said employees with a college degree, a lot of engineers and designers and so on. But by my estimate it supports almost 70,000 additional jobs in the metropolitan area outside Apple and outside high tech. [more to come, 52:14]

More EconTalk Episodes